This Project has three broad goals, each of which relate to the overall objective of this Program to develop novel therapies for hypoxic cells in tumors. A further integrating theme across all three goals is use of multicellular layer (MCL) cultures to quantify extravascular transport of drugs and their active metabolites. The first goal is to develop an improved analog of the hypoxia-selective cytotoxin tirapazamine (TPZ), which is currently in phase III clinical trial. The approach is based on insights from our spatially-resolved pharmacokinetic/pharmacodynamic (PK/PD) model for TPZ, which explicitly takes extravascular transport into account. This model demonstrates that large gains in activity could be achieved by increasing extravascular diffusion coefficients, optimizing rates of metabolism, and improving the fraction of metabolism that contributes to cytotoxicity (lambda). We will use the PK/PD model to guide advancement of compounds through screening, and will evaluate drug structure-activity relationships for each component of the model (e.g. diffusion coefficient in tissue). Our approach will test the specific hypotheses that (i) DNA targeting can be used to improve the cytotoxic potency and lambda, and (ii) increasing the intracellular diffusion range of the TPZ radical is a further strategy for increasing lambda. The second goal is to characterize bystander effects resulting from activation of prodrugs by enzymes expressed by recombinant clostridia in necrotic and hypoxic regions of tumors. This will involve the identification of active metabolites resulting from activation of prodrugs by E. coli nitroreductase, beta-glucuronidase and beta-galactosidase, and to quantify extravascular transport of the active species, and the resulting bystander killing, in MCLs. The third goal is to identify which of the initial "hits" from a screen for novel drugs targeting HIF-1alpha upregulation have good enough extravascular transport properties to warrant further development.